DESCRIPTION: (Applicant's Description) The absence of a comprehensive catalog of cytogenetic abnormalities, technical problems with sample preparation and heterogeneity of karyotype within a single tumor have hampered cytogenetic analysis of solid tumors as a clinical toot. New approaches to examine chromosomal changes from small samples of tumor or potentially even single cells, which scan a whole genome and are amendable to automation would be invaluable in cataloging karyotypic changes in clinical samples. Restriction maps can identify chromosomal rearrangements. When a map of a rearranged genome is aligned to a normal reference map, rearrangements become obvious. This proposal outlines the use of Optical Mapping to create a reference map of the whole human genome which will serve as the database by which chromosomal rearrangements associated with malignant cells are characterized at the molecular level in a model system. Optical Mapping is a robust system for the analysis of whole genomes. We have made whole genome maps of a number of microbial genomes. Our ability to mount and map very large human genomic DNA molecules encouraged us to constrict a map of the whole human genome. Optical Mapping techniques will be used to generate high resolution restriction maps from megabase-sized fragments prepared by simple shearing of DNA extracted from blood. We will explore further approaches to mounting large DNA molecules based on spermine condensation of metaphase chromosomes. Overlaps formed by alignment of ordered restriction maps derived from randomly cleaved genomic fragments will serve as the basis for the construction of whole genome maps without the need for library generation. A genome-wide ordered restriction map will be constructed with an average fragment I size of 25 kb using fragments 2 Mb-20 Mb. We project a whole genome map of high accuracy and minimal gaps will be obtained when 5 or more genome equivalents are mapped and contiged. However, significant sequence characterization of breakpoints will be accomplished before this goal is attained, using alignments obtained from 1-2 genome coverage maps constructed from both normal and cancer genomes. Our model system will be a lymphoblastoid cell line containing multiple chromosomal aberrations characterized extensively by cytogenetics. Ultimately, we expect to link these contigs and close a large portion of gaps by the Optical Mapping of BAC or YAC contigs previously assigned to the physical map. This will enable rapid and complete molecular chacterization of breakpoints and other genomic alterations. Finally, we propose to discern sequence information of those molecules characterized as harboring potentially breakpoint information. These approaches will entail molecular microdissection, and in situ PCR.